Gene therapy of brain tumors may be possible by using viral vectors to transfer toxic genes to tumor cells. A major impediment to such therapy will be the delivery of genetic vectors throughout the tumor and to invasive neoplastic cells distant from the main tumor mass. Transvascular delivery of virus by osmotic opening of the blood-brain barrier (BBB), may provide a means for global CNS vector delivery. During the previous funding period we continued our studies of virus and particle delivery to brain using osmotic opening, interstitial infusion, and pharmacologic manipulation of the blood-tumor barrier. We also demonstrated the presence of a physiological barrier at the level of the basement membrane, distal to the anatomic BBB, which may decrease the delivery of some virus-sized iron oxide particles to rat brain. Determining which properties allow passage through the basement membrane may have implications for increasing vector delivery. In Aims 1 and 2 of the current proposal, we will investigate the mechanisms of basement membrane binding of both iron oxide and virus particles in normal rats and in athymic rats with intracerebral carcinoma xenografts. We will assess the role of BBB charge, agent charge, and the dextran coating on iron particle delivery. The effect of tumor size and permeability, and the effect of altering the basement membrane with prior irradiation also will be evaluated. Delivery of iron particles across the anatomic BBB will be determined by magnetic resonance (MR) imaging, while delivery across the basement membrane will be assessed using light microscopy, and electron microscopy (EM). The delivery of AAV particles will be evaluated by fluorescence imaging of virus with capsids covalently labeled with the fluorophore Cy3. In Aim 3, a phase I clinical study will be performed using MR, histology, and EM to assess the abnormal permeability of the blood-tumor barrier to passage of iron oxide particles across the endothelium and basement membrane. Finally, in Aim 4, the relationship of delivery to efficacy will be explored. The Ad.TK recombinant replication defective adenovirus which is being used clinically by local injection bearing the herpes thymidine kinase gene will be tested for anti-tumor efficacy in combination with ganciclovir, after transvascular delivery to intracerebral tumor xenografts in the nude rat. Efficacy will be evaluated by MR tumor volumetrics and rat survival. We hypothesize that administration of virus with BBB opening will result in global delivery to brain tumors, providing enhanced efficacy. We anticipate that these studies will provide the basis for a clinical trial of osmotic BBB opening for brain tumor gene therapy.